I build the data pipelines and systems businesses actually run on.

What I do

• Data Engineering

  Pipelines that ingest, validate, transform and serve data reliably — scheduled, monitored, and documented.

  • ETL/ELT pipelines and bronze → silver → gold medallion architecture
  • Ingestion from 15+ web and system sources into one SQL-backed catalogue
  • Data quality checks and pipeline monitoring built into execution
  • SQL Server and PostgreSQL design — normalisation, stored procedures, indexing, tuning
  See the live pipelines →

• Development

  Production backends, APIs and interfaces — built, containerised, and deployed to cloud infrastructure.

  • Go REST APIs with JWT authentication over PostgreSQL
  • React frontends and Docker deployment (Railway, Cloudflare)
  • Bespoke workflow and approval automation, and Python scheduled jobs
  • Rules-as-data engines, append-only audit logs, and snapshot-based record design
  Read the Materials Requisition Portal →

• Systems Analysis

  Turning ambiguous, paper-and-spreadsheet processes into structured, auditable systems people actually adopt.

  • Requirements elicitation, process mapping and stakeholder facilitation
  • UAT leadership and Agile delivery across concurrent projects
  • Data flow diagrams, field mappings, runbooks and user guides
  • Approval, compliance and audit-trail workflow design
  Read the Compliance Workflow System →

Same pipeline as the panel above, in full: a scheduled Cloudflare Worker pulls UK grid carbon intensity every 30 minutes, lands it raw, transforms it through a bronze → silver → gold pipeline, and serves it from a documented public API. The chart below is real, not a screenshot.

• Materials Requisition Portal

  Full-stack Go/PostgreSQL/React system: alias-mapped catalogue import, full-text search, JWT role-based access and a snapshot-based requisition audit trail, deployed with Docker.

  • Go
  • PostgreSQL
  • React
  • REST APIs

  ↳ What used to be several people's personal spreadsheets became one searchable, versioned catalogue of real products — full-text search across thousands of items in place of manually scanning files, and a complete, point-in-time record of every requisition that doesn't quietly drift if the catalogue changes later.

• Supplier Data Ingestion Pipeline (15+ Sources)

  Python ingestion pipeline (requests, BeautifulSoup, Selenium, pandas) with isolated per-source extractors, normalising 15+ supplier sources into one SQL-backed catalogue.

  • Python
  • BeautifulSoup
  • Selenium
  • Pandas

  ↳ Product and pricing information across all supplier sources became queryable from one place instead of fifteen-plus separate manual lookups.

• Real-Time Asset Tracking Pipeline

  Event-sourced Python/SQL Server pipeline: QR scans ingested as immutable events, with current asset location and status derived from the event log.

  • Python
  • QR Codes
  • SQL Server
  • Pandas

  ↳ Asset location and status became something anyone could check directly, instead of something that depended on the spreadsheet being current.